Replacement of the Thiosugar of Auranofin with Iodide Enhances the Anticancer Potency in a Mouse Model of Ovarian Cancer

Replacement of the thiosugar of Auranofin with iodide enhances the

anticancer potency in a mouse model of ovarian cancer

Tiziano Marzo,

Lara Massai,

Alessandro Pratesi,

Matteo Stefanini,

Damiano Cirri,

Francesca

Magherini,

Matteo Becatti,

Ida Landini,

Stefania Nobili,

Enrico Mini,

Olivia Crociani,

Annarosa

Arcangeli,

_{Serena Pillozzi,}

_{* Tania Gamberi}

_{and Luigi Messori}

_*

a_{Department of Pharmacy, University of Pisa, via Bonanno Pisano 6, 56126, Pisa, Italy}

b_{Laboratory of Metals in Medicine (MetMed), Department of Chemistry “U. Schiff”, University of Florence, via della} Lastruccia 3, 50019 Sesto Fiorentino, Italy. E-mail: luigi.messori@unifi.it

c_{DI.V.A.L. Toscana s.r.l., via Madonna del Piano 6, 50019 Sesto Fiorentino, Italy}

d_{Department of Biochemical, Experimental and Clinical Sciences “Mario Serio”, University of Florence, viale G.B.} Morgagni 50, 50134 Firenze, Italy

e_{Department of Health Sciences, University of Florence, viale Pieraccini 6, 50139 Firenze, Italy}

f_{Department of Experimental and Clinical Medicine, University of Florence, viale G.B. Morgagni 50, 50134 Firenze, Italy.} E-mail: serena.pillozzi@unifi.it

KEYWORDS Auranofin; gold compounds; ovarian cancer; metal-based drugs; in vivo orthotopic model

ABSTRACT: In recent years, a few successful attempts were made to repurpose the clinically approved antiarthritic gold drug, Auranofin (AF), as an anticancer agent. The present study shows that the Iodido (triethylphosphine)gold(I) complex, (Et3PAuI

hereafter), – an AF analogue where the thiosugar ligand is simply replaced by one iodide ligand – manifests a solution chemistry resembling that of AF and exerts similar cytotoxic and proapoptotic effects on A2780 human ovarian cancer cells in vitro. Howev-er, when evaluated in a preclinical orthotopic model of ovarian cancHowev-er, Et3PAuI produces a far superior anticancer action than AF

inducing a nearly complete tumor remission. The highly promising in vivo performances here documented for Et3PAuI warrant its

further evaluation as a drug candidate for ovarian cancer treatment.

Auranofin, a linear gold(I) complex bearing triethyl-phosphine and thioglucose tetraacetate as ligands, is an orally administered drug, in clinical use since 1985 for the treatment of rheumatoid arthritis. AF is generally reputed as a reasona-bly safe drug for human use, even for long chronic treat-ments.1,2 _{Beyond its clinically established antiarthritic} proper-ties, in the frame of the so called drug repurposing strategies, AF has been extensively tested in recent years as a potential antiparasitic, antimicrobial and anticancer agent with encour-aging results.3-5 _{Of considerable interest are the} antiprolifera-tive properties of this compound that have been known since the early 80’s 5-7 _{and were later assessed in the NCI 60 cancer} cell line panel and in a few in vivo models.8-10 _{On the ground} of its anticancer properties, AF in the course of the last years has entered three distinct clinical trials, two of them (ovarian and lung cancer) still ongoing, and one (leukemia) closed.11-13 Results of these trials have not been disclosed yet. At the same time, considerable evidence has been gathered that a few spe-cific biomolecular targets such as the enzyme thioredoxin re-ductase, the proteasome system, the NF-κB pathway and other transcription factors are crucially involved in mediating the cellular effects of AF as well of other gold (I/III) com-pounds.14-18

It has been hypothesized, and then demonstrated, that the biological and pharmacological properties of AF might be kept and hopefully improved/modulated by replacing the thio-sugar ligand with other ligands such as various halides. In principle, replacement of the thiosugar ligand with halide lig-ands may lead to an increased lipophilicity and an enhanced drug bioavailability and biodistribution. On the other hand, this modification should not affect substantially the drug’s pharmacological profile, as the invariant [Et3PAu+] moiety is assumed to be the “true pharmacophore” whereas the thio-sugar ligand is believed to act mainly as a carrier ligand.18 Ac-cordingly, the actions of Iodido (triethylphosphine) gold(I) complex, Et3PAuI, and AF, have been comparatively

ana-lyzed toward ovarian cancer models both in vitro and in vivo. Et3PAuI is obtained by replacement of the thiosugar ligand

with an iodide ligand according to the reported synthetic pro-cedures18 _{(See ESI for detailed synthesis and characterization).} The chemical structures of AF and Et3PAuI are shown in

Figure 1. Chemical structures of Auranofin and Iodido

(tri-ethylphosphine) gold(I).

Replacement of the thiosugar ligand with iodide results into a greatly increased lipophilic character (a logP value of 4.6 has been measured by ICP-OES in the previous work18_{) with} re-spect to both Et3PAuCl and AF (this latter compound has

been previously described and used experimentally for rheu-matoid arthritis treatment),19 _{for which logP values of 1.7 and} 1.6 were measured, respectively.18 _{In spite of that, Et}

3PAuI is

still appreciably soluble in aqueous solutions where it mani-fests a very high stability as documented by a previous sys-tematic NMR analysis carried out in different environments including biological media.18,20 _{As reported in ESI (see figure} S4), new 31_{P NMR results clearly demonstrate a high stability} of Et3PAuI, with no changes in the structure of the molecule,

even in the presence of a large excess of sodium chloride (up to 100 mM). The documented stability in physiological-like conditions makes the compound an ideal candidate for biolog-ical testing.

The antiproliferative properties of Et3PAuI compared to AF

have been investigated in vitro toward A2780 human ovarian cancer cells. As displayed in table 1, both compounds produce potent cytotoxic effects with IC50 values falling in the high na-nomolar range. Et3PAuI is slightly more cytotoxic than AF

and both result significantly more effective than cisplatin, the well-known anticancer metal compound. These results are consistent with those previously obtained in colon cancer cell lines for the same gold compounds and recently published18 further supporting the concept that the presence of the thio-sugar ligand is not an essential requirement for the cytotoxic action and that substantial differences in the cellular uptake – despite the difference in the LogP value– between Et3PAuI

and AF are unlikely to occur. Remarkably, in agreement with a previous observation,18 _{no appreciable cytotoxic effects were} detected for Et3PAuI in normal cell lines i.e. human

fibro-blasts (HDF cell line) and human embryonic kidney cells (HEK293) over the concentration range 0−5000 nM.

IC50 (M) ±SD

Cell lines Et3PAuI Auranofin Cisplatin

A2780 0.394±0.174 0.754±0.139 1.53±0.66

n 4 4 5

p* 0.0138

SD, standard deviation; n, number of independent experiments; *t-test.

Table 1. Inhibitory effects of AF, Et3PAuI and cisplatin on A2780 cells following 72 hours drug exposure as determined by the sul-forhodamine B assay. Four independent experiments were per-formed.

Then, we examined whether Et3PAuI and AF are able to

induce apoptosis in A2780 cells; accordingly, we performed flow cytometry analysis of annexin V/propidium iodide-stained A2780 cells, treated with these two gold compounds, at the respective IC50 values, for 72 h. As shown in figure S1, the percentage of apoptotic cells in AF/ and Et3PAuI/ treated

samples is significantly increased (control vs AF, p=0.0121; control vs Et3PAuI, p=0.0107). In order to evaluate and

quan-titate apoptosis via caspase activity, we checked the activity of caspase 3 by FAM-FLICA Caspase-3 Assay Kit. The induc-tion of apoptosis observed for AF and its iodide analogue may be related with the strong thioredoxin reductase inhibition po-tency. Indeed, the IC50 values for the inhibition of the enzyme are generally in nice agreement with cytotoxicity values.18

The results collected evidenced that both gold compounds induce apoptosis through a caspase 3 activation, with AF be-ing slightly more effective than Et3PAuI (control vs AF,

p=0.0037; control vs Et3PAuI, p=0.04, see Supporting

Infor-mation).

Et3PAuI was subsequently evaluated in vivo in comparison

to AF in two different ovarian cancer models, i.e. a xenograft subcutaneous and an orthotopic model. Preliminarily, drug tolerability studies were carried out to assess the in vivo safety of Et3PAuI. The single-dose tolerability was evaluated in

athymic nude mice over a three weeks observation period. Et3PAuI was administered by intraperitoneal (ip) injection

using two different doses (20 mg/kg and 40 mg/kg). Only a single administration was performed, followed by a three weeks observation period; animals were controlled daily and no signs of sufferings such as weight loss, mobility reduction and appetite reduction were registered. In addition, a pilot study for gold biodistribution has been carried out and a gen-eralized higher accumulation for Et3PAuI, both in organs and

in the peripheral blood, was highlighted (see Supporting In-formation). Based on the above observations, we can hypothe-size that the detected differences in the in vivo activity of the two compounds could be due to their different lipophilic char-acter leading to an enhanced bioavailability and a more fa-vorable pharmacokinetics for Et3PAuI

The xenograft model consisted of 1x106 _{A2780 cells} im-planted subcutaneously (sc) in both flanks of the animals. The growth of the tumor masses was measured daily with caliper, and the treatment started when the volume of the masses reached 60 mm3_{. The treatment was performed by ip injections} (three times a week for two weeks of treatment) of AF and Et3PAuI (both 15 mg/kg). The growth of the tumor masses

was monitored over 21 days and time course is shown in fig-ure S3A and S3B respectively. At the end of the experiments, mice were sacrificed and the explanted masses measured; the volume (cm3_{) was calculated using the ellipsoid equation.}21 Remarkably, treatment with both Et3PAuI and AF was

ac-companied by a strong decrease of the tumor volume deter-mined at the sacrifice (figure 2, control: 2.96±0.06 vs Et3PAuI: 0.56±0.20, p<0.01; control: 2.96±0.06 vs AF: 0,79±0,27, p<0.01). The above experimental protocol did not alter the be-havior, vitality or weight of treated mice.

Figure 2. In vivo anticancer activity evaluation of gold

com-pounds in a xenograft subcutaneous (sc) model of ovarian cancer. Mean volume of tumor masses obtained at the sacrifice, after three weeks from injection of A2780 cells. Data are reported as the mean ± SD (*p<0.05, **p<0.01). Further details are reported in Supporting Information.

Subsequently, the in vivo anticancer activity of Et3PAuI

versus AF was assessed in an orthotopic model of ovarian cancer, a model that better represents the cancer behavior and response to therapy. Thirty female athymic nude mice were injected in the ovarian bursa (5x105 _{cells each mouse) with} A2780 cells transfected with the firefly luciferase gene (A2780-luc). Autopsy confirmed that ovarian cancer occurred in all animals and representative images of the procedure and of tumor masses are reported in figure 3A. The animals were divided in three groups of treatment (10 mice each): controls, AF treated and Et3PAuI treated. The administration schedule

was the following: mice received compounds (15 mg/kg) ip three times a week for two weeks. The rationale for such a treatment schedule relied on the schedule reported by Mirabel-li et al.5 _{and on the above reported biodistribution data} (Sup-porting Information) showing that administration of 15 mg/Kg gives rise to a peak peripheral blood concentration of 823.19 ng/ml and 944.17 ng/ml, for AF and Et3PAuI respectively,

after 24 hours. Mice were monitored daily, and the luciferase activity signal was quantified weekly (representative panels are reported in figure 3B).

Figure 3. (A) Orthotopic model establishment: injection of 1x106

A2780-luc cells in the ovary bursa of nude mice. Representative macroscopic images of ovarian tumor masses obtained are showed in the bottom panels (B) Representative pseudo-color BLI images tracking A2780-luc cells emission in mice, 7, 14 and 21 days after A2780-luc injection. Color bar represents light intensity levels reported as counts per minute (cpm). (C) Levels of ROI after 1 week of treatment, data are reported as mean of each group (n=10 animals/group) ± SD (**p<0.01). (D) Volume of tumor masses (mean ± SD) at the sacrifice after 3 weeks from injection of A2780-luc cells (control group, n=8; AF, n=5; Et3PAuI, n=5). (*p<0.05, **p<0.01). Further details are reported in Supplemen-tary Information.

Both gold compounds effectively reduced the tumor burden (figure 3B); notably, Et3PAuI is already very effective in

anti-tumor activity after few doses (i.e. 1 week of treatment) lead-ing to an almost complete tumor remission, as evidenced by the drastic reduction of the bioluminescent signal (ROI levels) compared to the control group (control: 20260±4631 vs Et3PAuI: 4729±2881; p<0,01) (see graph in figure 3C).

The tumor masses of the sacrificed animals after three weeks of treatment were collected and measured with caliper; the volume (cm3_{) was calculated using the ellipsoid equation}

even for this model.21 _{Most important, the tumor masses of the} animals treated with Et3PAuI showed a very strong reduction

of the volume compared both to the control group (control vs Et3PAuI p<0,05) and to the AF group (AF: 6.86±1.57 vs

Et3PAuI: 1.00±0.70, p<0.01) (figure 3D). This latter evidence

points out as Et3PAuI is a more effective anticancer drug

compared with AF in our animal model.

In conclusion, we have shown here that a specific structural modification of the repurposed gold(I) drug AF, now under consideration in clinical trials for cancer treatment, i.e. the substitution of the thiosugar ligand with iodide, affords a dis-tinct gold(I) complex Et3PAuI, whose anticancer profile in

vitro roughly resembles that of AF. This confirms that, even in the present ovarian cancer model, the thiosugar ligand is not mechanistically indispensable and that the [Et3PAu]+ moiety is likely to be the true pharmacophore. Very interestingly, in our in vivo models, Et3PAuI, has turned out to be far more

effec-tive than AF in contrasting tumor growth. Particularly impres-sive are the results and the differences between AF and Et3PAuI observed in the orthotopic model of ovarian cancer.

Et3PAuI is found to induce a very large tumor regression after

only one week of treatment, as pointed out by the decrease of the bioluminescent signal. A conspicuous shrinkage of the tu-mor is still evident after three weeks. The enhanced anticancer activity observed in this orthotopic model might well arise from the far greater lipophilicity of Et3PAuI, that might lead

to a more favourable pharmacokinetics and biodistribution. Notably, the improved in vivo anticancer activity of Et3PAuI

over AF does not bring about any increased systemic toxicity making Et3PAuI a promising new drug candidate. In view of

its far superior efficacy in animal models, Et3PAuI merits a

more extensive preclinical testing as it might be a valuable al-ternative to AF in cancer clinical trials.

ASSOCIATED CONTENT

Supporting Information

Synthesis and purity of Et3PAuI, logP determinations, Apoptosis plots; Cell death and caspase activity; Cell growth inhibition stud-ies (Sulforhodamine B assay); Tolerability studstud-ies; Biodistribu-tion studies; In vivo anticancer activity evaluaBiodistribu-tion xenograft sub-cutaneous models plots; NMR stability.

AUTHOR INFORMATION

Corresponding Author *E-mail: luigi.messori@unifi.it *E-mail: serena.pillozzi@unifi.it Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manu-script. ‡These authors contributed equally.

ACKNOWLEDGMENT

This study was supported by contributions of Associazione Ital-iana per la Ricerca sul Cancro, Milan (AIRC-IG16049) and Isti-tuto Toscano Tumori (Messori- ITT-2015) to LMe. TM thanks AIRC-FIRC (Fondazione Italiana per la Ricerca sul Cancro, Mi-lan) 3-years Fellowship for Italy (Project Code: 18044) and Uni-versity of Pisa (PRA_2017_25). CIRCMSB and ente CRF are gratefully acknowledged.

Animal experiments have been approved and authorised in ac-cordance with the Italian law, reference code: 113/2016-PR 03/2016.

REFERENCES

(1) C. Bombardier, J. Ware, I.J. Russell, M. Larson, A. Chalmers, J.L. Read, Auranofin therapy and quality of life in patients with rheumatoid arthritis. Results of a multicenter trial, Am. J. Med. 81 (1986) 565-578.

(2) C.F. Shaw, Gold-Based Therapeutic Agents, Chem. Rev. 99 (1999) 2589-2600.

(3) M.I. Cassetta, T. Marzo, S. Fallani, A. Novelli, L. Messori, Drug repositioning: auranofin as a prospective antimicrobial agent for the treatment of severe staphylococcal infections, Biometals 27 (2014) 787-791.

(4) E.V. Capparelli, R. Bricker-Ford, M.J. Rogers, J.H. McKerrow, S.L. Reed, Phase I Clinical Trial Results of Auranofin, a Novel An-tiparasitic Agent, Antimicrob. Agents. Chemother. 61 (2017) e01947-2016.

(5) C.K. Mirabelli, R.K. Johnson, C.M. Sung, L. Faucette, K. Muirhead, S.T. Crooke, Evaluation of the in vivo antitumor activity and in vitro cytotoxic properties of auranofin, a coordinated gold compound, in murine tumor models, Cancer Res. 45 (1985) 32-39.

(6) T.M. Simon, D.H. Kunishima, G.J. Vibert, A. Lorber, Cellular antiproliferative action exerted by auranofin, J. Rheumatol. Suppl. 5 (1979) 91-97.

(7) C.K. Mirabelli, R.K. Johnson, D.T. Hill, L.F. Faucette, G.R. Girard, G.Y. Kuo, C.M. Sung, S.T. Crooke, Correlation of the in vitro cytotoxic and in vivo antitumor activities of gold(I) coordination complexes, J. Med. Chem. 29 (1986) 218-223.

(8) H. Guo-Xin, L. Pan-Pan, Z. Shengyi, Y. Mengqi, L. Jianwei, Y. Jing, H. Yumin, J. Wen-Qi, W. Shijun, H. Peng, Elimination of stem-like cancer cell side-population by auranofin through modulation of ROS and glycolysis, Cell Death & Disease 9 (2018) 89.

(9) M.V. Rios Perez, D. Roife, B.B. Dai, Y. Kang, X. Li, M. Pratt, J.B. Fleming, Auranofin to prevent progression of pancreatic ductal adenocarcinoma, J. Clinical Onc. 34 (2016) 236.

(10) J.L. Roh, H. Jang, E.H. Kim, D. Shin, Targeting of the Gluta-thione, Thioredoxin, and Nrf2 Antioxidant Systems in Head and Neck Cancer, Antioxid. Redox Signal 27 (2017) 106-114.

(11) Auranofin in Treating Patients With Recurrent Epithelial Ovarian, Primary Peritoneal, or Fallopian Tube Cancer, https://clinicaltrials.gov/ct2/show/NCT01747798

(12) Sirolimus and Auranofin in Treating Patients With Advanced or Recurrent Non-Small Cell Lung Cancer or Small Cell Lung Can-cer, https://clinicaltrials.gov/ct2/show/NCT01737502.

(13) Phase I and II Study of Auranofin in Chronic Lymphocytic Leukemia (CLL), https://clinicaltrials.gov/ct2/show/NCT01419691.

(14) V. Scalcon, A. Bindoli, M.P. Rigobello, Significance of the mitochondrial thioredoxin reductase in cancer cells: An update on role, targets and inhibitors, Free Radic. Biol. Med. 127 (2018) 62-79.

(15) F. Magherini, T. Fiaschi, E. Valocchia, M. Becatti, A. Pratesi, T. Marzo, L. Massai, C. Gabbiani, I. Landini, S. Nobili, E. Mini, L. Messori, A. Modesti, T. Gamberi, Antiproliferative effects of two gold(I)-N-heterocyclic carbene complexes in A2780 human ovarian cancer cells: a comparative proteomic study Oncotarget 9 (2018) 28042-28068.

(16) A. Pratesi, C. Gabbiani, E. Michelucci, M. Ginanneschi, A.M. Papini, R. Rubbiani, I. Ott, L. Messori, Insights on the mechanism of thioredoxin reductase inhibition by Gold N-heterocyclic carbene compounds using the synthetic linear Selenocysteine containing C-terminal peptide hTrxR(488-499): An ESI-MS investigation, J. Inorg. Biochem. 136 (2014) 161-169.

(17) N. Micale, T. Schirmeister, R. Ettari, M.A. Cinellu, L. Maiore, M. Serratrice, C. Gabbiani, L. Massai, L. Messori, Selected cytotoxic gold compounds cause significant inhibition of 20S proteasome cata-lytic activities, J. Inorg. Biochem. 11 (2014) 79-82.

(18) T. Marzo, D. Cirri, C. Gabbiani, T. Gamberi, F. Magherini, A. Pratesi, A. Guerri, T. Biver, F. Binacchi, M. Stefanini, A. Arcangeli, L. Messori, Auranofin, Et3PAuCl, and Et3PAuI Are Highly Cytotoxic on Colorectal Cancer Cells: A Chemical and Biological Study, ACS Med. Chem. Lett. 8 (2017) 997-1001.

(19) B.M. Sutton, E. Mc Gusty, D.T. Walz, M.J. Di Martino, Oral gold. Antiarthritic properties of alkylphosphinegold coordination complexes, J. Med. Chem. 15 (1972) 1905-1908.

(20) T. Marzo, D. Cirri, S. Pollini, M. Prato, S. Fallani, M.I. Cas-setta, A. Novelli, G.M. Rossolini, L. Messori, Auranofin and its Ana-logues Show Potent Antimicrobial Activity Covering Multiresistant

Pathogens: Structure-Activity Relationships, Chem. Med. Chem. 13 (2018) 2448-2454.

(21) O. Crociani, F. Zanieri, S. Pillozzi, E. Lastraioli, M. Stefanini, A. Fiore, A. Fortunato, M. D'Amico, M. Masselli, E. De Lorenzo, L. Gasparoli, M. Chiu, O. Bussolati, A. Becchetti, A. Arcangeli, hERG1 channels modulate integrin signaling to trigger angiogenesis and tu-mor progression in colorectal cancer, Sci. Rep. 3 (2013) 3308.